Skip to content
KernelIndex
Search⌘K

ConvNeXtV2 block with GRN

22 eligible runs
other

ConvNextV2 block with Global Response Normalization (GRN). Combines depthwise 7x7 convolution, LayerNorm, pointwise expansion, GELU activation, GRN layer, pointwise projection, and residual connection. GRN computes L2 norm across spatial dimensions and normalizes by channel-wise mean.

Fastest reported
99.5µsmean · 62% of SOL
RIAC as wellLicense unknown · unknown

Reported evidence · last observed 2026-08-17. Reported by source; not independently reproduced. No public source. License unknown.

No public sourceRun detail →

Current records

Workloadsuite of 16 cases · mean latencypadding = 3 · kernel_size = 716 cases

Not measured on H100 for this workload. Challenges →

Source-native comparison · GPU NVIDIA B200 · Workload suite of 16 cases · mean latency · Protocol SOL-ExecBench evaluation stack v1.1 · mean · 16 results · last observed 2026-08-19Record history →
#
Implementation
Latency
vs #1
Trust
Observed
1
99.5µs62% SOL
1.00×
Reported · license unknown · no source
2026-08-17

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
2
100.0µs62% SOL
1.00×
Reported · license unknown · no source
2026-07-29

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
3
103.5µs61% SOL
1.04×
Reported · license unknown · no source
2026-08-15

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
4
109.7µs60% SOL
1.10×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
5
115.0µs58% SOL
1.16×
Reported · license unknown · no source
2026-08-04

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
6
117.5µs58% SOL
1.18×
Reported · license unknown · no source
2026-07-17

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
7
118.2µs58% SOL
1.19×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
8
119.2µs57% SOL
1.20×
Reported · license unknown · no source
2026-07-29

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
9
129.3µs55% SOL
1.30×
Reported · license unknown · no source
2026-07-27

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
10
139.0µs54% SOL
1.40×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
11
148.9µs52% SOL
1.50×
Reported · license unknown · no source
2026-08-19

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
12
149.6µs52% SOL
1.50×
Reported · license unknown · no source
2026-08-13

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
13
164.7µs49% SOL
1.65×
Reported · license unknown · no source
2026-08-07

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
14
172.3µs48% SOL
1.73×
Reported · license unknown · no source
2026-07-10

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
15
197.7µs45% SOL
1.99×
Reported · license unknown · no source
2026-07-11

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →
16
358.5µs31% SOL
3.60×
Reported · license unknown · no source
2026-07-25

Measured exactly what you asked. Reported by source; not independently reproduced.

no sourcelicense unknownno install recipeRun detail →

Implementations

Implementation
Runtime
Best latency
Evidence
Availability
unknown
99.5µs
1.00×
Reported
license unknown · no source
unknown
115.0µs
1.16×
Reported
license unknown · no source
unknown
117.3µs
1.18×
Reported
license unknown · no source
unknown
117.5µs
1.18×
Reported
license unknown · no source
unknown
129.3µs
1.30×
Reported
license unknown · no source
unknown
148.9µs
1.50×
Reported
license unknown · no source
unknown
169.8µs
1.71×
Reported
license unknown · no source
unknown
109.7µs
1.10×
Reported
license unknown · no source
Show all 22 implementations ›
unknown
118.2µs
1.19×
Reported
license unknown · no source
unknown
119.2µs
1.20×
Reported
license unknown · no source
unknown
135.5µs
1.36×
Reported
license unknown · no source
unknown
139.0µs
1.40×
Reported
license unknown · no source
unknown
149.6µs
1.50×
Reported
license unknown · no source
unknown
154.2µs
1.55×
Reported
license unknown · no source
unknown
164.7µs
1.65×
Reported
license unknown · no source
unknown
172.3µs
1.73×
Reported
license unknown · no source
unknown
197.7µs
1.99×
Reported
license unknown · no source
unknown
277.7µs
2.79×
Reported
license unknown · no source
unknown
358.5µs
3.60×
Reported
license unknown · no source
unknown
100.0µs
1.00×
Reported
license unknown · no source
unknown
103.5µs
1.04×
Reported
license unknown · no source
unknown
107.3µs
1.08×
Reported
license unknown · no source

Semantics

Inputs and outputs
xfp32 [batch_size, dim, height, width]
dwconv_weightfp32 [dim, 1, kernel_size, kernel_size]
dwconv_biasfp32 [dim]
layernorm_weightfp32 [dim]
layernorm_biasfp32 [dim]
pwconv1_weightfp32 [expansion_dim, dim]
pwconv1_biasfp32 [expansion_dim]
grn_weightfp32 [1, 1, 1, expansion_dim]
grn_biasfp32 [1, 1, 1, expansion_dim]
pwconv2_weightfp32 [dim, expansion_dim]
pwconv2_biasfp32 [dim]
epsfp32 scalar
layer_norm_epsfp32 scalar
outputfp32 [batch_size, dim, height, width]
Axes and behavior
dimvariable
widthvariable
heightvariable
paddingconstant = 3
batch_sizevariable
kernel_sizeconstant = 7
expansion_dimexpansion_dim = dim * 4
determinismunspecified
constraintsNo mutation or aliasing
Identity
alias035_convnextv2_block_with_grnmodelconvnextv2-base-22k-224sha256d7254e90066f…
Sources: NVIDIA SOL-ExecBench (2026-08-19)last observed 2026-08-19How records are decidedJSON